Characteristics of Underground Water Flow at Different Water Levels in Tianshengan Karst Area, Yunna

Vol. 84 No. 1 pp.206 -212ACTA GEOLOGICA SINICA (English Edition)Feb. 2010Characteristics of Underground Water Flow at Different Water Levels inTianshengan Karst Area, Yunnan, ChinaJanja KOGOVSEK*Karst Research Institute ZRC SAZU, Titov trg 2, SI-6230 Postojna, SlovenijaAbstract: Three tracing tests from the same injetion point executed at low, medium, and high waterlevels in the karst aquifer near Tianshengan village, Lunan Stone Forest, Yunnan Province, China, haverevealed the basic properties of underground water flow. They showed the general directions of waterflows; tracer concentrations were observed at six sccessive points allowing for the calculation ofapparent dominant flow velocities at these sections towards the Dalongtan karst spring. For the highwater level, the discharge between single sections was between two and 10 times greater than that at lowwater level. For the medium water level, the flow velocity at different sections was between 1.4 and 3.7times faster than that at low water level; and for high water level, it was between 1.3 and 2.7 times fasterthan that at medium water level. The fastest water flow appeared at the first section (23 cm/s at mediumwater level); and the slowest (0.6 cm/s at low water level) appeared where water flow must cross theTianshengan fault (north- south direction), and later, a layer of 20- 30 m thickness of quartz sandstoneand shale clay-stones. It was also possible to calculate the recovery of the tracer for point 4, Dakenyan,where discharge was measured. At the medium water level, 50% of the injected tracer was detected ahalf-day after its first appearance and at low water level after more than 3 days. The previouslypublished research ilustrates the transport velocities of possible contaminants and their solubilities inwater at different hydrological conditions.Key words: karst, tracing test, underground water flow, Lunan, Yunnan, China1 IntroductionNumerous research methods are used to establish thebackground of karst springs and the mode of water flowWater is the most important raw material in the world. Inthrough the aquifer. Natural tracing as detailed monitoringkarst areas, the rainfall directly disappears through theof physical parameters with data loggers of karst spring cancarbonate rocks and feeds karst aquifers. This watergive us more information about the recharge area of thereappears at the surface in karst springs. In karst areas, theyspring. Data on flow directions, velocities, and portions ofare the most important sources of drinking water. Inwater flowing in particular directions can be successfullyEurope, karst covers approximately 35% of the surfaceobtained by tracing experiments using artificial tracers(Biondic and Bakalowicz, 1995), in Yunnan 28% (Huangunder varying hydrological conditions.and Liu, 1998), and in Slovenia 43%. In Slovenia, moreTracing tests with artificial tracers have been proven tothan half of the drinking water comes from karst.be very useful and were systematically used within theThroughout the world, karst aquifers are becomingsphere of the Association of Tracer Hydrology more thanincreasingly important.40 years ago. The direction of water flow from the injectionBoth the quantity and also the quality of water arepoint, and in particular the velocity of discharge, dependsimportant. To preserve a suitable quality of drinking wateron hydrological conditions. Tracing tests from the samein karst areas, one must know the background of karstinjection point under various hydrological conditions showsprings well. With suitable planning of human activities inthe中国煤化工-locity of flow, and,such areas, a long -term warranty for water quality can beconsz/elocity of eventualprovided.polluMYHCNMHGofGegraphyfromKunming, Yunnan Province, China and the Karst Research* Corresponding author. E-mail; kogovsck@zrc sazu.siInstitute ZRC SAZU, Postojna, Slovenia have sucessfullyVol. 84 No.1ACTA GEOLOGICA SINICA (English Edition)Feb. 2010207worked on common projects. The Slovene andTLinkoupuChinese Ministry of Science and TechnologyTiansbengLinkoupu8a。made these feasible. Thus, the studies of theChinawater flow in Baiyun Cave show theaodoversaturated water due to the intensiveCM umn Soe ForeatOPoint 0SCS由_Smdissolution of carbonate rocks (Sebela et al,Point 62001). One of the study polygons of thecommon Chinese- Slovene project was also thekarst area around the village Tianshengan,OPointwhich is located in the northeasterm part of theT ianshenganLunan County, Yunnan Province, China.086Point 7082 Research AreaDakenyanooPointoPoint 4Point 8In order to determine underground water flow12@Point5。Dalongtancharacteristics in the studied karst aquifer inTianshengan area, a tracing test was proposed”ζ8oPoint'915Beidacun(Fig. 1). There is no surface drainage in the area;underground karst channels are well developed口17and are not deep below the surface. From pointdto the karst spring Dalongtan, that is, point 5Point 10(Fig. 2), the water flow is acessible through Fig. 1. Hydrogeological map (Afer Kogovsek et al..1997).several shafts and caves. The underground waterPoints 0-10. sampling points; 1, Precambrian non-cabonate rocks; 2, Carboniferous andPermian carbonate rocks; 3, Lower Permian clasic rocks; 4, fault; 5, cave or shaft with un-flow is as far as 8.9 km away from the Dalongtanderground water flow (sampling point); 6, cave or shaft with underground water nlow; 7, karstspring. The discharge of water flow is measuredspring (sampling point): 8. small spring; 9. surface stream: 10, lake or water reservoir, 11,in the underground channel at the Dakenyanweather station; 12, proven direction of underyround water flow based on tracing test inNovember 1998; 13, proven direction of underground water flow based on tracing tcst in Julypumping station (Fig. 3). At Dakenyan, the 1996; 14, pssible drecion of underground waler flow; I5, road; 16, vlag; 17, Shin (StoneForesl).precipitation stationmeasures dailyprecipitation.The climate of the studied area can be classified asDalongtan spring (point 5). We assumed that water atsubtropical with an average annual precipitation of 796points 6 (Wayaodong) and 7 (Guanyindong) and springs 9mm, average annual humidity of 75.3%, and an average(Changshuitang) and 10 (Xiniutang) had a different origin,annual temperature 15.6°C (1980- 1992). The amount ofbut we did not exclude the probability of connectiontotal annual rainfall varies considerably from year to year.(Kogovsek et al, 1997; Kogovsek, 1998).In the dry season, from October to April, when there is lessWe performed a combined tracing experiment fromthan 20% of total annual precipitation, the need for waterpoint d (Fig.4) and point 6 (Fig.5) at different hydrologicalirrigation is urgent. From August 1993 to August 1996, theconditions in the area; first at medium water level and latermaximal discharge of the underground water flow atat high and low water levels.Dakenyan was 6.28 m'/s and the mean discharge was 0.466In the tracing test at medium water level, the tracers werem/s (Kogovsek et al, 1997).injected on 19 July 1996. Uranin, a synonym of Sodiumfluorescein (Kass, 1998) was injected at point d (in water3 Used Methodsflow with the estimated discharge around 30 Ls) and NaClat point 6 (discharge 5 Ls only). At the time of injection theOn the basis of the first physical measurementsdischarge at Dakenyan was 0.783 m'/s and decreasing.(temperature, conductivity, and pH) and chemical analysesAfter rain on 22 July the discharge increased to 0.9 m/s(carbonates, calcium, magnesium, and CaMg ratio) atand then decreased slowly until the end of July (Kogovsekaccessible points, we assumed that the water flows fromet al., 1997).point d (Qinghuadong) towards point 0 (Yanshidong) andNoverThe中国煤化工was carried out innot in a direction towards point C, as had been supposed-)8 at point d, Uraninearlier. The analyses suggested that the water flows fromwas|YHC N M H Gwith discharge ofpoint 0 to point 1 (Maoshuidong), point 2 (Shihuiyao),approximately 30 40 Ls. The discharge of undergroundpoint 3 (Xiangshuidong), and point 4 (Dakenyan) up to the water flow at Dakenyan was at that time 0.246 m/s. In the208Underground Water Flow at Different Water Levels in Tianshengan Karst Area, YunnanKogovsekFig. 2. Dalongan karst spring (point5).Fig. 3. Discharge of water flow was measured in theunderground channel at the Dakenyan (point 4).Fig. 5. Injection of Uranin on September 1997 at Wayaodong(point 6).七C-mediumwl.。Clowwl4030200.100。Time after ijection (h)Fig. 4. lnjection of sodium chloride on September 1997 (point d).Fig. 6. Discharge and tracer breakthrough at Dakenyan (point4)atm中国煤化工，Tracervalues (Com). Normal-isedvalla); M is amount of tracerin mg:Hpresenlation of Cam insted of C (concentrations) has some advantage forcomparition of diferent tracers with dffrent injection amounts (Behrens ttal. 1992).Vol.84No.1ACTA GEOLOGICA SINICA (English Edition)Feb. 2010209following 10 days, discharge evenly decreased and on 2Table 1 Velocities of underground water flow at differentDecember it reached 0.156 m'/s (Kogovsek and Liu Hong,hydrological conditions at successive sections. Dominant1999). Fig. 6 clearly shows the bhydrological situation at thevelocities (Vdom) are the result of tracing tests with Uranintime of appearance of the tracer at Dakenyan (point 4) atand sodium chloride (designated with NaCI)Rclation Distance Vam VanVns/ VanpVlow and medium water levels.low medium_ highVeu__ VnsThe tracing test at high water level was performed on 26m_cm/s_ cm/scm/sSeptember 1997 when NaCl and Uranin were injected atd-0503018i5point d and point 6, respectively. During the injection, the1050.612500.2.2:; 2.2%6M0;5.8.7discharge at point 6 was found to be approximately 15 LJs,18500.57and near 100 Ls at point d. At the time of injection, the92450.4:3.1Nudischarge at Dakenyan was 1.38 m/s and increasing. It3350gYeGareached its maximum of 2.24 m'/s the next day (KogovSck46001 1964501.02 .3.029& Liu Hong, 2000).89020004 Results of Tracing Tests at Different45043505.9Hydrological Conditions15003.1NaThe tracing test in July 1997 at medium water level000土酒土适工楼二河γ16shows the water flows from point d in the direction ofpoints 0, I, 2, 3, and 4 to point 5. The fastest was water flowfrom the injection point d to point 0 (23 cm/s). Later, theoflow velocity slowly decreased (Table 1). The slowest wasthe water flow between point 2 and point 4. At the section喜100from point 2 to point 3 water flow must cross Tianshengan。Jfault (north-south direction) and from point 3 to point 4 the50|water crosses a less permeable layer of quartz sandstoneand shale claystones 20 30 m thick (Fig. 1). The average47%6 25.7.96 26.96velocity between injection point d and Dalongtan springwas 3.1 cm/s. Half day after the first appearance of tracer atFig. 7. Tacer breakthrough of the Uranin injected at point d andpoint 4 - Dakenyan almost 50% of injected tracer wereNaCl ijected at point 6 through sccessive points 2. 3. and 4t0recorded.the final Dalongtan spring (point 5) at medium water level.Conmn is concentration of uranin (ppb).At point 6 injected sodium chloride showed the mainwater flow towards point 2 and towards Dalongtan springpermanently active channels where water flows to the pointas already asessed by Uranin injected at point d and4 over points 2 and 3. However, I do not exclude theuncertain connections to points 7 and 9.possibility of mistake; this is why it would be reasonable toThe dominant flow velocity up to point 2 was 3.3 cm/s;double check the flow during the high water level. Waterthe velocities to the points 3, 4, and 5 were practically thetracing by Uranin showed a poor connection ofsame as those given by Uranin. This shows that in certainunderground water flow from point 6 to point 10, whilehydrological conditions these two water flows are united atconnection to point 9 was questionable and may bethe point 2 and continue as a uniform flow towards points 3neglected.and 4 up to Dalongtan spring, point 5 (Fig.7).The velocities of water drainage at low water level aren September 1997 at high water level, the tracer wasconsiderably lower than that at high water level. The fastestextremely diluted allowing us to follow the water waveis the water flow in the initial part (3.3 cm/s from point d tofrom the point d, where NaCl was injected, to the point 3point 2) and the slowest at the section between points 2 andonly, as further on the concentrations were too low and not4 (only 0.6 cm/s), similarly as we stated for water drainagereliable due to dilution.at medium water level. At point 4 the tracer recovery wasWater tracing by Uranin from the point 6 showed higher40% in 3 days after its first appearance. Unfortunately, atvelocities of water flow to the points 2, 3 and Dalongtanthat time. we finished the samnline as we did not expectspring in comparison with tracing test at medium watersuch a中国煤化工level. Tracer appeared at the point 4 earlier than the points 20nlTHCNMHGeIwereusedfortheand 3. This indicates the possibility of a direct connectioncompanson oI uralage velocues a particular sections ofbetween points 6 and 4 by the conduits lying higher thanthe underground flow during various water levels.210Underground Water Flow at Different Water Levels in Tianshengan Karst Area, YunnanKogovsekTable 2 Calculated maximal (Vmx) and dominant (Vom)oV-low_ ■V-medium ■V-highvelocities of some recent tracing experiments in Slovene10 1karst with injection in water fnow at sw alow-holes8Point of ingection Appearance ofDistance Vnax VamLracer in the springskm)(cm/s) (cm/s)_Vrtice6.02.3Ponikve - Mimag. K. Poljane10.0.6m■0.3Miklejev studenec Doblie0.6>oKrupa12RoEni dolKnupa8.05.4.2Bajer.9Ponikve - Gorjanci Metl Obh-2 2-344-5 d-3d-4 6-2-3-5.0SectionTezka vodaLokva . Predjama VipavaFig. 8. Calculated dominant velocities of underground waterTriscica - TenteraTominev studencc 21.1flow at different sections at low, medium, and high water levels.Javornikov izvir20.9The disances of sectios are in Table 1.DebeJakov izvir21.45 Comparison of Water Flow Velocities atthe medium water level showed that since the firstDifferent Hydrological Conditionsappearance of the tracer at point 4, one-half of the injectedtracer passed in the first 12 hours. The transport of UraninThe determined apparent dominant velocities 0started already before the increase of discharge. The fastunderground water flow in the Tianshengan area rangeand concentrated passage was controlled by precipitationsfrom 0.6 to 9 cm/s. Only in the initial part (d- 0) the flow isimmediately after the injection that increased the dischargefaster as it was recorded during the medium water level(from 0.48 to 0.9 m'/s) and pushed the water with the tracerwhen it was 23 cm/s. Flow at the high water level was theFig. 6). The further rinsing of the remaining Uraninfastest. At medium water level, the flow velocity at followed after every rain, that is, after every increase indifferent sections was between 1.4 and 3.7 times faster thandischarge, but this rinsing was less intense and lasted longerthat at low water level; and at high water, the level wasthan our observations.between 1.3 and 2.7 times faster than that at medium waterThe water tracing test at low water level shows a morelevel.flattened and lengthened tracing curve, which started toTable 1 and Fig. 8 show relatively fast drainage in theshape at point 4 after a longer temporal lag after theinitial section, d- -2, almost three times faster during theinjection (Fig. 6). The appearance of Uranin started at pointhigh water level than that at low water level. The flow4 at a low, stable discharge of 0.156 m/L. In this case, in 3slowed down the most between point 2 and point 4. Thedays only 40% of the injected tracer passed point 4.lowest velocity (0.6 cm/s) was recorded at the low water However, the transport of Uranin lasted longer even whenlevel. The comparison of flow velocities at variouswe did not sample any more. The retention times of Uraninhydrological conditions showed that at medium water level,at low water level and lower flow velocities are ratherthe water velocity at sections 2- 3 and 3- 4 is 3.5 times fasterlonger than that at higher water level.than that at low water level, and at high water level, it wasEstablished water flow velocities correspond to valuesan additional 2.7 times faster. Fig. 9 shows a fastgiven for underground waters of southem China by Yuanconsecutive appearance of the tracer at consecutive pointsDaoxian et al. (1991) and Song Linhua et al. (1993) citing2, 3, and 4 at medium water level and a substantial termporalvalues of several mm/s to more tens of cm/s.lag when water tracing was done during low water level.It can be concluded that the transport of eventualDuring medium water level most of injected tracer passedpollution is rather fast, especially during high and higherpoint 4 only 3 days after injection; while during low waterwater levels.level, the passage took more than 8 days.At the section from point 6 to point 2 the tracer was 1.36 Comparison with the Underground Flowtimes faster than that at medium water level. On the entireVelocities in Slovenian Karstwater flow from point 6 to the Dalongtan spring (point 5),the water flow was 1.9 times faster. As maximal dischargeSlovenia in the last 20is substantially higher than it was in the time of wateryears,中国煤化工tracer outlow wastracing although during high water level, we presume thatperfor个HC N M H Qn dropped below thedrainage in such conditions may be considerably faster.threshola oI aelecron, ana later atter nheavy rainfall (HabicThe calculation of the returned tracer at point 4 duringet al. 1990; Habic & Kogovsek 1992; Kogovsek & PetritVol.84 No.1ACTA GEOLOGICA SINCA (English Edition)Feb. 2010211100faster than that at low water level; and at high water level itwas between 1.3 and 2.7 times faster than that at medium75water level. The comparison of stated velocities of karstwater drainage in the Tianshengan karst aquifer and inSlovene karst shows that in the Slovene karst the flow25velocities are lower.The transport of the tracer at medium water level is500substantially faster than that at low water level. At mediumTime afer injection (b)water level, 50% of the injected tracer was detected atFig. 9. Comparison of tracer breakthrough at sccessive pointsDakenyan in 12 hours after the first appearance of the2, 3, and 4 along the underground water flow at medium andtracer. At low water level, the transport of Uranin waslow water levels.slower, 40% of tracer passed the point 4 in more than 3Tracer concentrations are expressed as normalized values (Com).days.2002，2006). Obtained results enabled accurateThe data of drainage velocities are useful as they showdetermination of Vmax (at the first appearance of the tracer)how fast the eventual contaminants would spread atand Vdom (at the maximal concentration of the tracer). Waterdifferent hydrological conditions through the treatedtracing tests in the Slovene karst, when the tracers wereaquifer.injected directly into the water flow, indicated dominantflow velocities up to 2.4 cm/s, and 4.6 cm/s at a higherAcknowledgementswater level (Table 2). At low water level, the velocities inthe Slovene karst are roughly four times lower than that atI would like to thank the Yunnan Institute of Geograpby,high water level.the Slovene Ministry of Higher Education, Science andTechnology and the Chinese Ministry of Science and7 ConclusionsTechnology who enabled the research work. I am gratefulto Maja Kranjc, Karst Research Institute ZRC SAZU forThe first tracing test in the karst aquifer nearthe translation of my text.Tianshengan village near the Lunan Stone Forest, Yunnan,China at medium water level with inection at point dManuscript received Sept. 17, 2007showed the direction of underground water flow towardsaccepted Oct. 6, 2008Yanshidong (point 0), Maoshuidong (point 1), Shihuiyaoedited by Zhang Xinyuan(point 2), Xiangshuidong (point 3), Dakenyan (point 4), upto the Dalonglan spring (point 5). This underground flow isReferences8.9 km long. At Wayaodong (point 6), the injected sodiumBehrens, H Benischke, R.. Bricelj, M, Harum, T.. Kass, Kosi,chloride showed the main water flow towards ShihuiyaoG., Leditzky, H.P.. Leibundgut, Ch., Maloszewski, P.. Maurin,V., Rajner, V., Rank, V.. Reichert, B.. Stadler, H., Stichler, w..(point 2) and towards Dalongtan spring, as already assessedTrimbom, P., Zojer, H., and Zupan, M, 1992. Investigationsby the Uranin injected at point d (Fig. 1).with Natural and Atificial Tracers in the Karst Aquifer of theAn average flow velocity at the medium water level toLurbach System (Peggau-Tanneben Semriach, Austria).point 2 from point d was 4.7 cm/s, and that from point 6 wasSteirische Beirdige zur Hydrologie, 43: 9-158.3.3 cm/s. The slowest is the underground water flow fromBiondic, B. and Bakalowicz, M..1995. Hydrogeological aspectspoint 2 to point 4 where water flow must cross theof groundwater protection in karstic areas. In: COST Action65. Final Report, 3-7.Tianshengan fault (north- -south direction), and later a layerHabic, P.. Kogovsek, J.. Bricelj, M., and Zupan, M.,. 1990.of 20-30 m thickness of quartz sandstone and shaleDoblicica springs and their wider karst background. Actaclaystones (2.1 cm/s at medium water level). FurtherCarsologica, 19: 5-100.tracing tests at low and high water levels with the sameHabit, P.. and KogovSek, J.1992. Water tracing in Krupa karstcatchment, SE Slovenia. Acta Carsologica, 21: 35-76.injection points gave the basic properties of undergroundHuang Chuxing and Liu Hong, 1998. Karst of Yurnan. In: Southwater flow at different hydrological characteristics.China Karst. Ljubljana: Zbirka ZRC, 19: 11-17.The determined apparent dominant velocities range fromKass, W.. 1998. Tracer Technique in Geohydrology. A.A.0.6 to 9 cm/s; only at the initial part the water flow wasconsiderably faster. At high water level, the discharge atKogov中国煤化工, 1997. Properties ofsingle sections was between two and 10 times faster thanund 1near Lunan in YunnanrouYH.CN.M H GTracer Hrogy97:that at low water level. At medium water level, the flowProceedingsnf the 7th Intemational Symposium on Watervelocity at different sections was between 1.4 and 3.7 timesTracing, Portoroz Slovenia, 26- 31 May 1997. Rotterdam: A.212Underground Water Flow at Different Water Levels in Tianshengan Karst Area, YunnanKogovSckA. Balkema, 255- -261.landfill near Ribnica in southeastem Slovenia. Acta Carsol,Kogovsck, J, 1998. Physical and chemical characteristics of35(2): 91-101.groundwater of Tianshengan area. In: South China Karst.Sebela. s., Slabe, T., Kogovsek, J. Liu Hong and Pruner, P.,.Ljubljana: Zbirka, Z.R.C.. 19: 91-98.2001. Baiyun Cave in Naigu Shilin, Yunnan Karst, China. ActaKogovsek, J. and Liu Hong, 999. Water tracing test in theGeologica Sinica (Engl. ed.), 75(3): 279- -287.Tianshengan region, China at low water level in NovemberSong Linhua, He Yueben and Feng Yan, 1993. Ground water1998. Acta Carsologica, 28(2): 241-253.tracing in Wulichong surface drainage system Mengzi county,Kogovsek, J. and Liu Hong, 2000. Water tracing test in theYunnan province. Proceedings of the 1Ith IlnternationalTianshengan region, Yunnan China at high water level. ActaCongress of Speleology, Beijing, 227 -229.Carsologica, 29(2): 249- -260.Yuan Daoxian, Zhu Dehao, Weng Jintao, Zhu Xuewen, HanKogovsek, J. and Petrit, M..2002. Underground water flow fromXingrui, Wang Xunyi, Cai Guihong, Zhu Yuanfeng, Cuithe Trziscica sinking stream (SE Slovenia). Acta Carsologica,Guangzhong and Deng Ziqiang， 1991. Karst of China.31(2): 75 -91.Geological Publishing House, Beiing, 1- 224.Kogovsek, J.. and Petric, M., 2006. Tracer test on the Mala gora中国煤化工MYHCNMHG